Seven Signals is a free 104-page field manual for reading startup momentum in public GitHub activity. It explains seven repeatable signals, includes a 90-minute replication workflow, and opens the supporting 55-startup, 5-quarter dataset (219 startup-period observations). Read online or download PDF, EPUB, Markdown, and plain text. CC BY 4.0, with SSRN methodology, a Zenodo dataset, and a free MCP server for Claude and Cursor.
Hey Product Hunt,
I wrote Seven Signals because fundraising announcements are late. The engineering work starts earlier, and it leaves public traces.
The book turns those traces into seven checks you can run yourself: commit-velocity acceleration, contributor influx, infrastructure buildout, star detachment, issue-close-time, repository expansion, and founder technical presence.
It is free to read online and free to download as PDF, EPUB, Markdown, or plain text. The supporting panel is open too: 55 venture-backed startups over five quarters, 219 startup-period observations, published on Zenodo under CC BY 4.0. The formal methodology is on SSRN.
I also included a 90-minute replication appendix and a free MCP server, so Claude or Cursor can query the live signal feed.
The honest caveat: this is a screening method, not an investment oracle. The strongest statistical result is an association on a small labeled subset. The point of publishing the book and data is to make the method easy to test, criticize, and improve.
I would especially value feedback on leakage and look-ahead bias in event-prediction work.